Abstract
Accurately predicting gas transport in rocks is required for enhancing the accuracy of field production models. The mesoscale lattice Boltzmann (LB) method can be implemented to predict gas permeability in porous rocks. However, the published LB results for the Klinkenberg effect are often inconsistent with the widely used Beskok-Karniadakis-Civan's (BKC's) correlation. The culprit of the unphysical effect has been identified in the typically implemented boundary conditions (BCs). An improved BC is proposed herein to reliably predict gas permeability. Non-equilibrium molecular dynamics simulations are conducted to benchmark the proposed approach. The results show that the presented LB predictions for the Klinkenberg effect are quantitatively consistent with experimental data and the BKC's correlation, indicating that the unphysical effects have been minimized. More importantly, a numerical consistency is achieved for describing the Klinkenberg effect at molecular through macroscopic scales. These observations are relevant for improving our ability to predict gas production from tight formations.